PHTI finds virtual CKD care has not delivered the kidney benefit investors expected
Billions have flowed into virtual chronic kidney disease management, but the clinical return remains hard to prove. A new evaluation from the Peterson Health Technology Institute found no consistent evidence that eight leading virtual CKD platforms slow kidney-function decline or materially reduce overall healthcare spending.
The findings carry weight because CKD is not a niche condition. Roughly 37 million Americans live with the disease, and 87% do not know they have it. The gap between prevalence and diagnosis leaves a large population outside the care pathways that digital vendors often promise to improve. For clinicians, the central question is not whether these platforms can send reminders or generate engagement. It is whether they change the biological course of progressive nephron loss, albuminuria, and the downstream march toward dialysis and cardiovascular complications.
PHTI reviewed eight prominent virtual CKD management solutions and drew on more than 5,400 articles and other evidence sources. The breadth of that review matters because digital health vendors often accumulate testimonials, utilization data, and selective performance metrics long before hard outcomes are available. Here, the signal was weak. Across the products evaluated, there was no consistent evidence of slower kidney-function decline and no consistent increase in use of guideline-recommended medications that are already known to protect kidney health, including SGLT2 inhibitors and nonsteroidal mineralocorticoid receptor antagonists.
The absence of a durable prescribing effect is especially important. In CKD, the drugs with the strongest disease-modifying logic act on pathways that drive ongoing injury: intraglomerular pressure, sodium handling, tubular stress, and inflammatory and fibrotic signaling. If a virtual platform cannot reliably move patients onto those therapies, its clinical influence is likely to remain superficial, even when the user experience appears polished and the interface keeps patients engaged.
One measured benefit did emerge, and it was narrow. Some patients using virtual CKD services began dialysis in a planned setting rather than through an emergency “crash” start. But the effect reached only about 1 in every 1,000 patients with CKD. In a condition as common and expensive as CKD, a benefit that small may help individual patients, yet it is too limited to justify broad claims of population-level transformation.
The spending analysis was similarly restrained. For a 1-million-member Medicare Advantage plan, the companies evaluated manage more than $5 billion in annual healthcare spending, but the reduction in spending was only about 0.1%. In practical terms, that is not the kind of cost signal that changes benefit design, payer strategy, or employer purchasing behavior. It also suggests that virtual care efficiency, as currently deployed in CKD, does not automatically translate into fewer hospitalizations, fewer late-stage complications, or lower downstream renal replacement costs.
Emergency dialysis starts improved, but the signal was too small to anchor a business case
The emergency dialysis finding deserves attention because it points to a plausible clinical mechanism. Earlier identification of advanced CKD, closer surveillance of worsening renal function, and better care coordination can help some patients avoid unplanned dialysis initiation, a scenario often associated with catheter starts, urgent hospitalization, and avoidable morbidity. Even so, the size of the effect was tiny relative to the burden of disease. A program that helps a small fraction of patients avoid a crash start may improve individual care trajectories, but it does not amount to broad disease modification.
This is where many digital kidney-care models appear to fall short. They may improve communication, streamline follow-up, or organize data, yet those gains do not automatically alter the specific clinical behaviors that determine renal survival. Early diagnosis, blood pressure control, consistent albuminuria monitoring, and uptake of renoprotective medications remain the interventions most likely to slow CKD progression. If a platform is not moving those levers, its downstream impact will remain limited regardless of how frequently patients log in or respond to prompts.
The scale of the disease burden makes that limitation difficult to ignore. Traditional Medicare spends approximately $141 billion annually on people with CKD, while employers face an estimated $107 billion in direct costs tied to the disease. Those figures help explain why the virtual CKD sector drew so much capital in the first place. They also show why payers and purchasers now want evidence that reaches beyond utilization metrics. Engagement is not the same as disease control, and download counts do not substitute for better kidney outcomes.
PHTI’s recommendation is blunt: contracts and incentive programs should be redesigned around patient-level clinical milestones rather than engagement or utilization metrics that have weak links to outcomes. In CKD, that means rewarding earlier diagnosis, uptake of guideline-recommended medications, and hard clinical endpoints instead of counting app use, outreach frequency, or message response rates. The logic is straightforward. A platform that changes the clinical trajectory should be able to demonstrate movement on biologically meaningful markers, not just member activity.
For nephrologists and primary care physicians, the report reflects a familiar frustration. The technology stack has expanded faster than the evidence base supporting it. Digital tools can help organize care, but CKD progression is still governed by pathophysiology, not software adoption. Glomerular hyperfiltration, progressive fibrosis, and vascular risk do not respond to engagement dashboards unless those dashboards lead to actual treatment changes.
The evaluation also places pressure on investors and health systems that have treated virtual CKD care as a scalable answer to a costly chronic disease. The current data do not support that assumption. The platforms reviewed did not consistently slow decline in kidney function, did not consistently increase use of proven therapies, and produced only a minimal spending effect. In a field where the stakes are kidney failure, dialysis, and cardiovascular risk, those are not small omissions.
The remaining challenge is not whether virtual CKD care can create convenience. It can. The harder question is whether any of these models can reliably alter the clinical course of disease at scale. PHTI’s evaluation suggests that, so far, the answer has not been demonstrated.